Can you bet the server farm on HP?

One year later a post-merger Hewlett-Packard is still tilling the soil to prove it made the right technology decisions

Carly Fiorina, Hewlett-Packard’s chairman and CEO, chose her words carefully. “The new HP will be a company you can count on,” she declared.

Her proclamation went forth during Fiorina’s keynote address at HP World Conference & Expo in September, only four months after HP completed its $19.7 billion acquisition of Compaq one year ago this week.

The speech exemplified Fiorina’s mission to rewrite the enterprise IT history books. It used to be said that no one got fired for buying IBM, the company upon which HP has squarely trained its crosshairs since the merger. But can the same be said for HP? With the passage of time, technologists are seeking reassurance that HP has backed the right products in the merger.

One of the four executives who planned and oversaw the merger, Shane Robison argues that HP got every technology decision right. “I can’t think of an instance where if we could start over again, we would do it differently,” says Robison, executive vice president and chief strategy and technical officer at HP in Palo Alto, Calif.

Every decision was made with the understanding that HP must build products and services based on a lower-cost model, Robinson explains. “Our value proposition is to deliver high-technology solutions, which is a combination of services, software, and hardware at a best-in-class price. If you compare that with [competitors] Dell and IBM, our approach is to be high tech, low cost.”

The old HP way manifested as a slow decision-making culture with technically elegant but expensive products. What has emerged is an aggressive, street-fighting style driven by a conviction that competitively priced products with a clear technical advantage will give HP the edge in the consolidating IT industry. Also driving HP’s new approach is a desire to project the image of a comprehensive enterprise services provider on par with IBM.

But according to analysts, the road ahead will not be easy. “You have to give [Fiorina] credit for bringing on capable management, but they don’t have a strong direction of distinguishing advantage. IBM has them beat in terms of being the worthy, reliable enterprise provider. Dell has them beat in terms of cost and can provide all the same HP can, with the exception of the legacy systems. And Sun has a strong Unix presence,” says Will Zachmann, president of Canopus Research in Duxbury, Mass.

HP’s Robison and other HP executives counter that their value proposition holds up under competitive pressure. IBM is too expensive, and Dell lacks product breadth and a services business.

“We are certainly less proprietary in the overall solution stack than IBM, which allows us to continue to push hard on the price side of things,” Robison says.

Most industry observers believe HP has given itself a fighting chance to succeed — with but a few exceptions — in its decisions regarding which products would survive the merger. Good technology decisions include retaining Compaq’s ProLiant line of Wintel-based servers rather than HP’s Wintel lineup, such as the NetServer.

One user who had relied on the NetServer product had no problems migrating to Compaq’s ProLiant product. “We haven’t seen any degradation of products or services. We did see the obsolescence of some HP server lines, but we expected some to go away,” says Ken Parchinski, CTO of Automated Resources Group in Montvale, N.J.

Other thumbs-up from observers were given to the decision to retain Compaq’s Storage Works products and the accompanying ENSA (Enterprise Network Systems Architecture). Since moving forward with Compaq’s lineup, HP has jumped into the top spot for overall disk-based storage, according to IDC.

HP officials attribute the improvement to the ENSA framework and a focus on midsize companies in the face of declining enterprise-storage business. “The combined HP has about 40 percent market share of the modular, midrange market. We will continue to make investments there for management, more efficient resource utilization, and a range of different virtualization technologies. We will do better integration with OpenView or things like virtualizing the datacenter,” says Howard Elias, executive vice president in charge of HP’s storage products.

Another area generally met with approval has been HP’s bold move to abandon the development of middleware technology acquired from the purchase of Bluestone. HP executives dumped the integration software, citing the technology’s increasing commodity status of middleware, instead focusing on bolstering its systems-management solution OpenView.

According to HP, OpenView will also mature as a platform that supports the company’s UDC (Utility Data Center) strategy, a set of server-based technologies HP is developing to deliver utility-based computing.

“What we will continue to do is marry the intelligence of OpenView with the ability to reallocate a datacenter with the UDC, as well as keep investing heavily in the Web services management area,” says Nora Denzel, senior vice president of HP Software. “This should give us the ability to provide business-process management in addition to the infrastructure management and service management you have today.”

The datacenter plans reflect what HP’s Robison said was HP’s ability to support selected platforms. The company supports three operating systems with Windows NT, Linux, and HP-UX, and is partnering with BEA, Microsoft, and Oracle for the missing middleware pieces.

HP is counting on enterprises needing a company besides IBM, Microsoft, or Sun to support the complexity of integrated .Net and J2EE environments. To that end, shortly after the merger, HP formed an independent software group inside the company that, according to Denzel, has the feel and function of an ISV.

“This ISV inside the company will help as we double down on our efforts surrounding management and the manageability of everything from business process to the services they are linked into regardless of what hardware you are using,” Denzel says.

Meanwhile, perhaps the best nondecision HP has made was to make HP’s lucrative printer and imaging products standard. Compaq has never had or wanted a meaningful presence in those markets.

Tough Calls

Yet not every decision has been perceived as a clear-cut winner. HP’s decision to back Intel’s 64-bit Itanium 2 microprocessor across its entire server line remains controversial. IBM and Dell declined Intel’s partnership offers, leaving HP to stand with Microsoft as the two major companies to support Itanium 2.

This decision is an example of what Denzel calls “focused innovation.” With HP-UX, Windows, and Linux all capable of running on Itanium-based systems, users need only dial up whichever operating system they need for any given server or workstation, thereby simplifying deployment and management issues.

Although Itanium is generally regarded as a powerful chip, major issues have delayed its adoption, including the difficulty of implementing a brand new instruction set, a lack of support for

32-bit applications, and a lack of significant ISV support.

In addition, HP must now contend with the launch of AMD’s 64-bit Opteron, which promises interoperability with 32-bit apps. “When you come up from the [32-bit] Xeon space, Itanium looks attractive. But when you come down from [Compaq’s] Alpha [chip], you have serious concerns about the power for the dollars,” says Hal Kuff, a vice president of technology and an HP user at wireless technology distributor Tessco Technologies in Hunt Valley, Md.

Longtime HP user Kees den Hartigh, a systems analyst at The University of Alberta in Edmonton is also wary, concerned about Itanium’s capability of replacing the PA-RISC architecture. “[Itanium is] cheaper to produce with help from someone else, but putting all your eggs in one basket is a dangerous thing to do. PA-RISC is still a very powerful chip and a lot of people have a vested interest in it,” he says.

Regardless of the concerns, Robison said HP expects to release a full set of platforms based on the latest version of the Itanium chip, code-named Madison, by the end of this year. The company’s goal is to extend Itanium “across the entire datacenter, from the Web tier through the application tier to the database tier,” he says.

PC Turf

Another decision still provoking debate is the maintenance of two separate desktop PC lines. In the minds of many, HP still has too much overlap between products in this hotly contested space.

HP has pursued a dual consumer-brand strategy, marketing the Compaq Presario and HP Pavilion desktops. HP chose to make the Pavilion into its high-end, consumer desktop machine, whereas the Presario line offered a step down in available features, says Jim McDonnell, vice president of sales and marketing in the personal systems group at HP.

The result has seen HP jockeying with Dell for the glory of No. 1 PC maker, a game of PC-leapfrog analysts believe could last at least through the end of this year.

Meanwhile, HP faces the risky proposition of supporting both Linux and Microsoft. “Linux has created a new tension in this whole ball game. Three or four years ago, OEMs placed their bets on Microsoft or Microsoft. But Linux has created a new power dynamic, and I think HP needs to play the Linux card very carefully and shrewdly,” says Dana Gardner, a vice president at Boston-based The Yankee Group.

Services to the Rescue

By the same token, HP has already started to prove it can bridge competing platforms from a solutions-provider standpoint, supporting what ultimately suggests the entire merger bet is paying off.

HP signed a 10-year, $3 billion managed-services contract with Proctor & Gamble last month. On the same day, it also landed a lucrative outsourcing deal with Ericsson, proving that large enterprises are willing to bet the farm on the new HP. Company officials attribute the wins to the new HP’s size and capability of offering one-stop shopping.

One impressed customer is Grant Westcott, head of technology infrastructure at the Canadian Imperial Bank of Commerce (CIBC) in Toronto.

CIBC awarded a seven-year, $1.5 billion deal to HP in September 2002. HP is still learning how to get the most effective synergy from its merged services unit, but the potential for great results is evident, CIBC’s Westcott says. “I can’t give you a sound bite saying ‘Yes, it’s perfect.’ But I can say that it’s getting more and more interesting by the day,” he says.

HP’s Robison notes that services and outsourcing deals will increasingly make up the majority of HP’s revenues. And when it comes to finding a stable services company, revenues speak volumes: HP generated revenues of $16.5 billion, $18.0 billion, and $17.9 billion during its first three quarters as a combined company. “What we need to do [now] is get the other parts of the business executing equally well,” Robison said.

Tom Krazit and Juan Carlos Perez contributed to this report. They are correspondents at IDG News Service, an InfoWorld affiliate.